Working Paper: CEPR ID: DP7685
Authors: Alfonso Irarrazabal; Andreas Moxnes; Luca David Opromolla
Abstract: When trade costs are of the iceberg type (Samuelson 1952) and markups are independent of trade costs, relative prices across markets are distorted, but relative prices within markets are not. When trade costs depart from the analytically convenient iceberg type, distortion will also occur within markets. In this paper we build a heterogeneous firm model of trade that allows for both iceberg and per-unit costs. An important theoretical finding is that these within-market distortions create an additional channel of gains from trade through within-industry reallocation. We fit the model to firm-level export data, by product and destination, using a novel minimum distance estimator and find that average per-unit costs, expressed relative to the consumer price, are 35-45%, depending on the elasticity of substitution. The pure iceberg model is therefore rejected. Finally, we calibrate the model and quantify the costs of protectionism. Simulations indicate that the welfare costs are roughly 50% higher when tariffs are per-unit compared to when they are iceberg.
Keywords: exports; intraindustry reallocation; trade costs; trade liberalization
JEL Codes: F10
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
Cause | Effect |
---|---|
trade costs (F19) | relative prices across markets (F16) |
trade costs (F19) | relative prices within markets (P22) |
per-unit trade costs (F12) | relative prices across markets (F16) |
per-unit trade costs (F12) | relative prices within markets (P22) |
per-unit trade costs (F12) | intra-industry reallocation (L16) |
intra-industry reallocation (L16) | welfare (I38) |
per-unit trade costs (F12) | consumer prices (P22) |
per-unit tariffs (L97) | welfare costs of protectionism (D69) |